population, Gordon and Wyon (18, 19) have described their findings obtained during long-term studies of population dynamics in villages in the Punjab of India. They followed the 1,479 children born in their study villages from 1955 to 1958 and thus could calculate accurate mortality rates in these children. When mortality rates were correlated with maternal parity, which they accepted as a sufficient indicator of family size, they obtained the results summarized in Table 3. TABLE 3 Mortality of 1,479 Children Born in Eleven Punjab Villages, by Parity of Mothers, India, 1955-1958 Parity of mother 1 2 3 4 5 6 7-12 Unknown Totals Number of births 230 209 210 197 165 136 326 6 1,479 Neonatal mortality (deaths/1,000 infants up to 28 days) 95.7 52.6 81.0 30.5 84.8 51.5 95.1 166.7 73.7 Infant mortality (deaths/1,000 infants up to lyear) 171.8 116.5 144.9 123.7 171.8 164.2 206.3 166.7 160.6 2nd-year mortality (deaths/1,000 pop.) 75.8 15.6 24.2 92.4 95.7 76.9 95.0 0.0 67.9 Source: Wyon and Gordon (19). Their figures show clearly that mortality in those infants tended to increase with family size. Among second- or later-born children the trend is most apparent, though in the first month of life, before family or environmental circumstances are as likely to affect the newborn, differences are not great. After 1 month they are striking, and it is important to note that the effect is relatively much greater in the second year of life. Mortality rates for the first year among seventh- or later-born children are not quite twice as high as those in second-born children in the first year of life; in the second year they are over six times as great. Similar observations have been made in several countries at the other end of the development scale. Some of the most impressive evidence to be found concerning the association between family size and infant mortality conies from the studies of Morris, Heady, Morrison, and their associates in a study of all the births that occurred in England and Wales during 1949 and 1950. They matched the information on the death certificates of all children dying under as the probability of picking one from a small family. In this situation, one would conclude that average size of family is 9.2 children. . . .